The current development version of
hesim now provides a general framework for integrating statistical models with economic evaluation. Users can build a decision model by specifying a model structure, which consists of a set of statistical models for disease progression, utility values, and costs. Each statistical model is used to simulate outcomes as a function of estimated parameters. N-state partitioned survival models (PSMs) and individual-level continuous time state transition models (CTSTMs) are now supported.
Fitted statistical models
formulaobjects. These include
params_create objects storing samples of parameters of fitted statistical models for probabilistic sensitivity analysis.
create_params()is a generic function for creating parameter objects from a fitted statistical model or a
formulaobject. Parameters can be sampled using Monte Carlo multivariate normal approximations or via bootstrapping.
params_surv_list()) and linear regression (
params_lm()). Splines and parametric distributions (exponential, Weibull, Gompertz, gamma, lognormal log-logistic, generalized gamma) are supported for survival modeling.
hesim_data()creates an object of class
hesim_datafor storing a collection of data tables or data frames for simulation modeling.
expand.hesim_data()combines some or all of the data tables or data frames in
hesim_data()into a single long dataset.
input_data()creates an object of class
input_data, which contains data for predicting or simulating values with a statistical model.
create_input_data()creates an object of class
input_datafrom a fitted statistical model or a
Health state values
StateValssimulates the costs or utilities associated with health states.
StateValsobject from fitted statistical models or
Partitioned survival models
Psmsimulates outcomes from N-state PSMs.
Psmobject is instantiated with a set of survival models (the R6 class
PsmCurves) and models for costs and utility (the R6 class
PsmCurvesobject from fitted statistical models or
Continuous time state transition models
IndivCtstmsimulates individual-level CTSTMs. Semi-Markov (i.e., “clock reset”) models are currently supported.
IndivCtstmobject is instantiated with a health state transition model (the R6 class
CtstmTrans) and models for costs and utility (the R6 class
CtstmTransobject from fitted statistical models or
psm4_exdataprovides example datasets for parameterizing a PSM.
ctstm3_exdataprovides example datasets for parameterizing a CTSTM.